ADVERTISEMENTS
Researchers invent a technique that could cram 10 trillion bits onto a surface the size of a quarter.For decades, hard-disk makers have been shrinking the electromagnetic media that store data to cram more information into a tinier footprint. Now two nanotechnology researchers may have shown the next step in that ever-smaller evolution: Letting the disks make themselves. Ting Xu, a professor at the University of California, Berkeley and Thomas Russell, a professor at the University of Massachusetts at Amherst, have created a technique that could, theoretically, pack a disk the size of a quarter with 10.5 terabits (more than 10 trillion bits) of data, the equivalent of 250 DVDs.
Source: Forbes
Source: Forbes
The secret to packing that much information on such small real estate–about 15 times denser than the densest data storage device currently in existence–is self-assembly, or tricking the disk’s materials into organizing into an array of data-storing dots packed far tighter than what could be accomplished with current techniques.
Nanotechnology techniques such as self-assembling promise researchers an intriguing alternative to continuing to refine the most commonly used technique for building transistors on silicon computer chips, namely, optical lithography. Optical lithography techniques become increasingly problematic as engineers build tinier devices–the wavelength of the light used to cast patterns onto silicon is already bigger than the width of the devices engineers want to create.
That’s forced them to use ever more exotic tricks to shrink the wavelength of light. (See “Intel’s Stimulus Plan.”) That means making chips with yet tinier features–and keeping pace with Moore’s Law dictum of doubling the density of transistors on a chip every two years or so–threatens to become excruciatingly difficult.
“If you can’t keep up with Moore’s Law, forget it,” says Russell. “This is beating Moore’s Law by a couple orders of magnitude.”
Here’s how they do it. Russell and Xu’s method starts with a sliced crystal, either sapphire or silicon, cut at an angle that exposes a ragged section of the crystal’s lattice structure. They then heated the crystal for 24 hours to up to 2,700 degrees Fahrenheit–a process that forces the crystal surface to reorganize itself into a sawtooth pattern at three-nanometer intervals (at that breadth, 30,000 such jagged edges side by side would be as wide as one human hair).
The researchers then spray the crystal’s surface with a specially designed polymer dissolved in a hydrocarbon-based solvent. After drying and being treated with another solvent, the polymer settles into a regular hexagonal pattern–a kind of plastic-like screen–on the crystalline surface. Vaporizing nickel onto the surface and removing that screen leaves behind a near-perfect grid of metal, hexagonal dots. Each of those dots, in theory, could hold a distinct magnetic state, representing a one or a zero.
Commercially manufacturing this kind of nano-hardware data storage still remains years, if not decades, away. One challenge would involve creating a magnetic head that could hover over those dots, read their signal and write new data onto the grid. Another barrier would be scaling the process so that the self-assembled nanoscale data grids could be stamped onto a surface automatically. Hal Rosen, a manager of research at Hitachi (nyse: HIT - news - people ), says he’s impressed by the researchers’ ability to shrink so much storage into such a small space, but skeptical that a self-assembling method could create larger-scale disks. “It’s a promising approach, and they’ve made some nice breakthroughs,” Rosen says. “But in this industry, we have incredibly severe requirements in terms of precision. The question is whether they can do this over a large area and be able to predict exactly the location of any single dot.”
Russell is optimistic, however, that their process could work at different size scales. The only limit to the total area of the disk, he says, is the area of the crystal slice that begins the process.
If nanoscale data storage became possible at the level Russell and Xu imagine, the result could be anything from enormously dense enterprise storage systems to tiny iPod-like devices with more than a terabyte of storage space. Getting to that stage might mean combining self-assembly methods with lithography, or doing away with the head that reads the data-packed disk and instead creating a self-assembling solid-state system on the model of Flash memory storage used today.
“I admit that there are several problems that still have to be solved,” says Russell. “But they’re solvable.”
0 comments